3

A tuner accuracy is ±0.5 cent. Does that mean that it is highly sensitive or just decently sensitive for pitch accuracy?

3

According to Wiki., a one cent change is imperceptible to most humans, so half of that is going to be a pretty accurate parameter change. One semitone (say from E to F, or A to Bb) is easily recognised, but when that interval of a minor second is split into 100 different 'notes', most wouldn't spot note 77 from 78 as being different/out of tune. Halve that difference, and I'd say it's certainly accurate enough for me! Just how much more accurate would one need it to be?

  • Just for reference, +/- 5 cents is considered the just noticeable difference which is where most people will start to hear a difference rather than the "same note" played one after another (although if two pitches played within 5 cents of each other beating may occur). – Dom Jun 24 '18 at 5:47
  • @Dom - listening to beating is how I tune. It's especially good on basses, where beats cycling by three or four seconds can still be noticed. What I don't know is, for example, what beats of say, 1 second might represent in terms of % accuracy, but it must depend on the frequency involved. – Tim Jun 24 '18 at 6:10
  • 2
    @Tim There is not a simple relationship between cents and beat frequency. A cent represents a fixed proportional between frequencies. A beat of 1Hz represents an absolute difference. So, a 1Hz beat between two low notes means the notes are quite far apart but a 1Hz between two high notes may be trivial. formulas.tutorvista.com/physics/… – badjohn Jun 1 at 9:14
  • Worst case scenario: the tuner puts A 440 a half cent flat and A 880 a half cent sharp, or vice versa. The two played together would beat with a period of about two seconds. But it seems rather more likely that whatever inaccuracy is present within the advertised range, it would probably apply to all pitches. – phoog Jun 2 at 20:23
1

I do my own tuning using TuneLab software. I try to get all strings within 0.5 cent of perfect. Then I check for beats, especially very slow ones, like 2-3 seconds. It sounds good when I play, so I must be doing something right. At the first sign of beating I touch up the tuning. (It helps to have a Dampp-Chaser installed.)

FWIW, I've been playing - and restoring - a 1943 Baldwin Acrosonic and have just upgraded to a 1959 Baldwin model M grand. The grand is in amazing condition for its age but for some reason was a few Hertz below A440. I'm about in the middle of tuning it, which is what prompted my search for what kind of tolerance is acceptable.

  • 1
    I thought that the naive mathematically exact tuning of pianos did not sound good and the extreme octaves are usually stretched. This is due to the harmonics of strings not being exact. en.m.wikipedia.org/wiki/Stretched_tuning – badjohn Jun 1 at 9:19
0

It depends on for what purpose. For tuning a violin or guitar, you should be fine. For tuning a rigid-pitch instrument with composite notes (like an accordion or organ) tuning two notes independently with 0.5 cent error each that are supposed to be sounding in unison (either with the same frequency or a fixed ratio like 2:1 or even the 3:2 of an organ's quint registers) is not going to sound well.

  • 1
    Isn't that effect used in most chorus pedals? – Tim Jun 24 '18 at 8:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.